Bacterial vaginosis (BV) is an exceedingly common disorder of the vaginal microflora affecting >30% of all women, with higher rates in pregnancy and among African- American populations. Women with BV are at substantially increased risk of preterm birth, which is a major cause of neonatal morbidity and mortality, as well as acquisition of sexually transmitted diseases including human immunodeficiency virus. Despite its public health importance, the pathogenesis of BV is not well understood. We have recently characterized vaginolysin (VLY), a cholesterol-dependent cytolysin from Gardnerella vaginalis (a bacterial species present on the vaginal mucosa in the setting of BV and thought to contribute to the pathogenesis of disease) that exhibits exquisite human specificity. We hypothesize that this species-specific toxin may be an important virulence factor of G. vaginalis with relevance to the pathogenesis of BV. In our preliminary data, we have characterized the receptor for VLY (human CD59) on genital tract epithelial cells. Introduction of this receptor into non-susceptible cells renders them sensitive to VLY. We have engineered a transgenic mouse expressing the hCD59 receptor and also constructed a VLY chimera that is hCD59-independent. These represent candidate in vivo models for BV. In addition, we have developed techniques for genetic manipulation of G. vaginalis, including transposon mutagenesis. In Aim 1, we will define genetic determinants of G. vaginalis virulence using new techniques for mutagenesis and assays of toxin production. In Aim 2, we will determine the role of VLY at the host-pathogen interface both in vitro and in vivo with a focus on unique aspects of the VLY-hCD59 interaction. At the conclusion of these studies, we will have expanded our knowledge of G. vaginalis pathogenesis, evaluated new in vivo models of BV, identified candidate strategies to inhibit toxin-host interaction, and developed new tools for continued investigation into the pathogenesis of an important disorder.